Psuedo Blocks in a Petrus speed solve.

[Cyrus C.]

Does anyone use Psuedo Blocks for Petrus? I know Kirjava (I think) sometimes does it for Roux, but does anyone do it for Petrus? I've been experimenting with it casually & the only problem I have is EO, & Corner recognition, which I could probably get the hang of with practice. Or is there an obvious reason for not using it that I'm missing?

 

[CharlieCooper]

I tried and found recognition absolutely impossible. Granted, the way it was introduced to me was by Arnaud as we were driving at about 70km down a Dutch motorway, but I still don't think it's for me. If you can nail recog though... HELL YEAH.

 

[Cyrus C.]

Well I think I may practice it for a bit, but I'm not going to use it in speed solves yet. Although, I may start trying for Psuedo Blocks that are just different in the M layer, the corner recognition is what I'm really having trouble on.

Try the scramble: S2 R U R' U R' U2 R2 y R U' R' L' U' L

You'll find a 2x2x3 Psuedo done, it's really easy to solve actually if the M layer is the only disturbed thing.

 

[Johannes91]

If there's a nice pseudo 2x2x3 (it's pseudo, not psuedo ), I'll take it and finish Heise style. Unless it uses all 6 colors, then I'll discard it. I haven't tried continuing with Petrus much but it feels really annoying and hard.

After a normal 2x2x3, I occasionally make a pseudo 1x2x3 block in Step 4. The ones that are off by R2 don't affect LL recognition much at all. Those that are off by a 90-degree turn are a bit more tricky, especially if you use ZBLL, but it's doable most of the time.

 

[Cyrus C.]

If there's a nice pseudo 2x2x3 (it's pseudo, not psuedo ), I'll take it and finish Heise style. Unless it uses all 6 colors, then I'll discard it. I haven't tried continuing with Petrus much but it feels really annoying and hard.

After a normal 2x2x3, I occasionally make a pseudo 1x2x3 block in Step 4. The ones that are off by R2 don't affect LL recognition much at all. Those that are off by a 90-degree turn are a bit more tricky, especially if you use ZBLL, but it's doable most of the time.

If you make a pseudo 1x2x3 block in step 4, wouldn't LL recognition be the same as making a pseudo block earlier?

 

[iasimp1997]

I think my friend does. But I have no idea.

 

[Kirjava]

Half turn pseudoblocks are easier than quarterturn. Having just opposite colours is a lot easier to recognise than random colours. However, quarterturns should also be easy if you know hyperorientations.

It's hard to practise pseudoblocks for speedsolving, since they only really happen by chance (no point in using it unless it's going to lower your movecount by much). Using one is a bit of a gamble.

I don't think pseudoblocks in the 2x2x3 would be worth it at all. Even opposite ones make the rest of the solve go crazy. Unless you're doing this, I don't see why EO recognition should be any different.

I don't think slicepseudoblocks are very smart, either. I can't imagine situation where they would be really worth it, as you don't get the advantage of a lot of pieces being solved easily by doing it. Also, your solves would end in a random slice move. Nobody likes that.

However, an opposite block in Step 4 shouldn't affect your recognition much at all. Not if you're doing OLL/PLL, anyway.

The reason why pseudoblocks are so useful with the method Gilles Roux describes is because as long as you're opposite neutral, it doesn't have an effect on your recognition at all. Also, you don't have to do an extra move to undo the solve conjugation - it's simply a case of doing r2 instead of M2 at the end of the solve.

EDIT: Johannes ninja'd me on a lot of these points >_>

If you make a pseudo 1x2x3 block in step 4, wouldn't LL recognition be the same as making a pseudo block earlier?

If you make a pseudoblock earlier, recognition for steps other than the LL suffer.

 

[Athefre]

One thing I've considered before is to solve the UL and UR edges in the DF/DB slots. If you're a CLL/ELL user, it would take a couple of moves off your overall average. In the end you would be left with a simple "M2", "UM2", or "U'M2" and an M-Slice permutation.

 

[qqwref]

I can see how pseudoblocks in a Petrus solve could be useful, but unless you're talking about doing step 4 an R2 off I really doubt there is anything to be gained in a speedsolve. Recognition can be quite difficult and you're only talking about a handful of moves; the loss of good lookahead and the necessity to think pretty much completely negates this benefit. Think of how many Fridrich solvers you see doing F2L with a pseudoblock. Doing this won't be even close to effective (by which I mean better than the solve would've been if you'd done it normally) without a ridiculous amount of practice, and I suggest just not bothering with the idea.